Skip to main content

DeepSeek · Reasoning · mai-ds-r1

MAI-DS-R1: Specs, GPU requirements, hosting & deployment

MAI-DS-R1 is a recent Reasoning release in the DeepSeek family, certified from its immutable official source revision.

Reference only Compare this model
Technical deployment reference illustration for MAI-DS-R1
Reference deployment flow; exact architecture facts are listed separately.
Total parametersThe manufacturer has not published this information.
Active parametersDense model
Context windowN/A / Not directly comparable
Release statusRecent · Recent
Architecture
ModalitiesText
Licensemit
Deployment stateReference only
Commercial stateCommercial Use Allowed
Release date

Architecture, strengths and use cases

MAI-DS-R1 is an official self-hostable DeepSeek release in the DeepSeek Reasoning track. It uses The manufacturer has not published this information., carries The manufacturer has not published this information., supports a verified context window of 0 tokens, and exposes only source-certified checkpoints and runtime combinations. Its release, license and deployment facts are tied to the official repository revision rather than inferred from the family name.

MAI-DS-R1 uses The manufacturer has not published this information. with The manufacturer has not published this information.. The stored architecture profile records layers, hidden width , attention heads and a 0 token context ceiling where the official configuration publishes those fields.

Verified The manufacturer has not published this information. architecture
Official checkpoint and pipeline evidence
Official Unknown checkpoint coverage
Private LLM inference
Agentic workflows
Long-context document processing

Certified checkpoints and quantizations

Only exact, source-certified variants attached to this release are shown.

MAI-DS-R1

microsoft/MAI-DS-R1

Official · Unknown · safetensors

Precision / quantizationFormatBitsPublisher trustDeployment state

Recommended hardware

Use only the listed certified runtime configuration with the exact checkpoint precision. The recommendation engine checks weight memory, KV cache, runtime overhead, safety headroom, GPU architecture and topology before ranking rentable and reference hardware.

Reference only

The official license permits commercial use; its obligations still apply.

License and self-hosting considerations

The full 0B resident weight set must fit across the selected GPUs. Context and concurrency add KV-cache memory; MoE active parameters do not replace resident-weight memory.

MAI-DS-R1 is published under mit. The release-scoped official license link and verification date are retained with the catalogue record.

Commercial Use Allowed. The official license permits commercial use; its obligations still apply.

Alternatives and sibling variants

DeepSeek-R1, DeepSeek-R1-0528, MAI-DS-R1-FP8, DeepSeek-R1-0528-NVFP4-v2, DeepSeek-R1-0528-NVFP4, DeepSeek-R1-NVFP4-v2, DeepSeek-R1-NVFP4, DeepSeek R1 0528, DeepSeek R1

Release timeline

deepseek/deepseek/mai-ds-r1

Source details

Official source

Source revision: 41a6bb82abd08a303092ea65ce98aaa094f727e017e54d66c730f9a3c5e743f3

Frequently asked questions

How much VRAM does MAI-DS-R1 require?

The exact estimate depends on checkpoint precision, context and concurrency; the page recommendation uses the certified memory engine and compatible runtime configurations.

Which runtimes support MAI-DS-R1?

No runtime is presented as supported until upstream evidence is verified.

Can MAI-DS-R1 be used commercially?

The stored commercial classification is Commercial Use Allowed. Consult the official mit terms for the final obligations.

Official references

Last verified: 2026-09-03